How he grew his AI startup from $2M to $20M ARR in 12 months. | Omar Haroun, Co-Founder of Eudia

I started to feel like we have Product Market Fit. Anytime anyone talks to one of our customers, we get another customer. So, like, our whole GTM engine is kind of how do we just, like, try to get our current customers in a room with prospects and at that point, like, we can walk away, which I think was a pretty good sign that we're onto something. Yeah, we hit a million ARR right around six months. We weren't focused on ARR, I would say, for the first couple of years, but in the last 12 months, we went from 2 to 20 million ARR. I think a lot more about how. How do we get to a billion in revenue by 2030 and build a generational company and win this market than I do. How do we, like, optimize for sales cycles and this year's revenue?

0:00

That's Product Market Fit.

0:40

Product Market Fit. Product Market Fit. I called it the Product Market Fit question. Product Market Fit.

0:42

Product Market Fit.

0:46

Product Market Fit.

0:47

Product Market Fit.

0:49

I mean, the name of the show is Product Market Fit.

0:50

Do you think the Product Market Fit show has Product Market Fit? Because if you do, then there's something you just have to do. You have to take out your phone, you have have to leave the show five stars. It lets us reach more founders and it lets us get better guests. Thank you. Omar, great to have you on the show, man.

0:52

Yeah, great to be here.

1:09

Looking forward to this one. You just raised a massive $100 million series a not too long ago, and the legal tech and AI and legal space is just, like, completely blowing up. We had Lagora on here, Max from Lagora recently we had Ryan from filevine. We've had Spellbook back in the day. Bluejay, I guess, is more on the tax side, where they're close by. In any case, you know, everybody seems to not just be raising money, but doing well and driving, like, pretty insane business. So maybe as a first question, like, you know, when you think about your Harvey Lagora Favine kind of thing, where. Where exactly does UDIA sit in?

1:10

Yeah, definitely. So our view has always been that AI is not the future of software. It's actually the future of labor. And if you look at certain buyers in the enterprise, and I would include the Chief Legal Officer in this category, but there's many more, by the way. The Chief Procurement Officer, the head of HR, even the CFO, typically 95 to 98% of their budget goes to humans. Labor services, not software. And so we basically asked ourselves from day one if historically there's been a lot of great software that's never really been adopted. It may be the case that no amount of AI on top of the 2% of the budget will ever really change anything. And so to really offer like the 10x better alternative to that 98% of the budget, we decided to really deeply look at what are the humans who consume that budget actually doing and how do we much more directly, first of all not sell to these law firms or professional services firms who have a lot of kind of perverse incentives. Right. They don't really want to cut hours because they bill by the hour and instead focus on the end customer who has shared incentives and think about building a Data platform and AI platform that can actually offer a 10x better alternative to what they're doing today.

1:45

So this is a good segue into something else that I also happen to believe, which is that there's a massive opportunity in AI enabled services. And you know, your choice with legal AI, really just about any type of AI is you can either sell it like software was sold, right? So you go to a law firm and you're like, here's rai. If you use it, you'll be more productive, more efficient, and then you can go and deliver whatever services you deliver to the end customer, or you can just say what it sounds like you're doing and you can correct me if I'm wrong, which is, you know what, we'll use the AI and we'll deliver these legal services to the end customer. And now you become a kind of a competitor to these law firms instead of, you know, a provider, a vendor to them.

2:58

Yeah, so. So it's definitely a part of what we're doing and I would say more something that we did because what I've seen for the in house enterprise legal team, which is where we've only ever sold, we haven't sold to law firms, is there's the highest standard deviation, but call it roughly 50% of the budget goes to law firms, 50% goes to the internal headcount. Right. And team. And so for the work that's already going out the door and leaving the building, it was actually often a better mechanism for us to just be able to not wait for law firms to change, but to just really demonstrate that a lot of the work they're paying a hundred thousand dollars for can now be done for a thousand dollars. And actually it's better, faster and cheaper, not just cheaper. So that was kind of like a good way for us to, I would say, get on a lot of people's radars. And very, very quickly prove the value for the work that's being done in house or that they want to be in house. We actually do provide software and it's, it's, it is an end to end data platform and so we typically have customers doing both. Like for some work they're more or less just shipping it to RAI agents and it's more of a UX is more like what they get with the law firm where they give an input and they get back an output. But then for probably the vast majority of again like the work that we do, it's actually more providing software for the teams internally.

3:38

And that like distinction is not unlike the other one. So we had like 10x which is like you know, AI for, for MSPS. Quanta which is AI, not AI from SPS and AI first MSP like themselves and then Quanta which is an AI first bookkeeper. It's a lot harder to go in like you're talking about in house legal versus the contracted work. The in house legal is a lot harder to replace and also there's a lot less incentive to replace that because not only is it a full time employee and you've got all the legal ramifications of that. A full time employee tends to have more. I don't know about complicated work but like work that just spreads in many different areas. It's rarely just a single task. A lot of the outsourced contracted stuff is just easier to switch providers. I mean you're already set up or you have an outside provider so to say, okay, now we're going to have a different provider has, you know, and at least in the way these other companies are doing it seems a lot more doable. My question to you though, just to fully understand it is are you equally focused on the in house legal teams and replacing the outside contract work or have you like shifted more towards one or the other?

4:55

Yeah, so I, I think it's, it's evolving because you know when you look at the history of in house legal teams, GE kind of pioneered this like 40 years ago and really the idea started out quite simply which is if I'm paying law firms like the equivalent of a million dollars a year to do M and A, or I could hire a FT for 200k, it's kind of a no brainer to get somebody internal for. So it was kind of. Actually the whole point of having an in house team is to be able to reduce the cost and reliance on law firms. I think what ended up happening is these chief legal officers Found that it's not just cheaper to actually have somebody who has all the context of the business. You're their only client. They start to add a lot of value because even if there's no M and A deal going on right now, they're paying attention to what's happening to business or figuring out ways to add value outside of just pure legal. Right. And so the reason I point that out is as I almost feel like our product roadmap took a similar path, which was like day one, if a customer has like an M and A deal and they, and we both know that the M and A due diligence is like 50 to 70% of the costs of the legal fees. And all that can now be like mostly automated, but not entirely automated. But unfortunately the law firms today make a ton of money off of doing it the old school way. We can come in and say, hey, great, we even have a law firm now. Our law firm can be co counsel to your law firm in like two weeks. We can prove that we can save you 500k, get you kind of what you need 10x faster than the law firm, whatever, do it. And it kind of proved the value. I think what ended up happening was people got so excited about what we did that they were like, I want to have this internally for my team as well. And so now I would say actually the vast majority of what we do is we're building this brain for each customer because how each customer thinks about risk is actually quite different. And so once we build the brain, which could be trained off of their external matters or internal matters or ideally both, and we start to understand how does this organization actually think about risk and then we're embedding that into their workflows. So basically to your point, it's, it might be like a longer onboarding process. We have to get access to all their internal data systems, all of that. But when we do it, and now that we've had, you know, two years to be at this, it's been incredible what the results are.

5:56

And so you're now, if I understand correctly, more focused on serving these in house legal teams and automating a lot of the stuff that they do versus kind of just becoming an AI first law firm and taking over the world that way.

8:07

Yeah, like the first 18 months, to be clear, we were actually only doing the internal stuff and then we created and launched the law firm and that was an accelerant. When you think about like the in house team, like let's just say 50% of the work is external and 50% is internal. They're not only frustrated that they're outsourcing their dollars to the law firm and the rates actually have gone up 18% year over year since AI came out. So it's weird because the law firm is putting out press releases about using AI and they're not really seeing the savings right directly. So that's like a frustration. But I think even more on a deeper level, they're not just outsourcing their dollars, they're actually outsourcing their knowledge. And so if you think about like the in house chief legal officer's perspective by analogy with AI, about like these human experts externally are using your company's training data to build their own brain and then renting that expertise back to you at a higher and higher hourly rate. So I think what we've really started to do is actually find a mechanism to not only take all of this institutional knowledge that we can glean from internal systems, but also what are all the learnings from you having used law firms over the years and even going forward and how do we actually bring that back to keep that knowledge internal?

8:18

So that was good, like context maybe. Let's go back to the beginning. Where, where were you, you started this company like late 2023. Where were you early 2023? Like what were you doing before?

9:26

Yeah, I was finishing up my time at Relativity. So I'd sold my last company, which is also an AI company focused on Fortune 500 in house legal and we were acquired by Relativity, which is a very, very large, probably $10 billion software company that's, that's dominating kind of in the litigation space. And then I led AI strategy there for a couple years and really, really enjoyed that experience. But yeah, I was, I was at the tail end of that time, maybe

9:35

without getting too deep into it. What did your last startup do and like how big did it get by the time it sold?

9:58

Yeah, it was pretty interesting cause we were doing again AI from 2014 to 2021, which is the one we got acquired. The old AI, it was exactly like the way that we built it was like teams of PhDs doing all the ML ops like data labeling and very, very painful stuff that's no longer necessary. But we focused on like a pretty interesting niche which was like one of my frameworks that I use sometimes for enterprise software selling is there's three ways to help a customer. You can help the customer make money, save money, or keep the CEO out of jail. We were doing the third thing with my last Company so literally RAI would help in like these large scale litigations. We'd say, hey, keep your armies of lawyers, don't even think about saving money or efficiency. Just give us the data and we'll flag. Here's like evidence of the CEOs having an affair. And none of this is relevant to the litigation, but you don't want to hand it over to the public record and have a disaster on your hands. And so we would find these like privileged and sensitive documents in these like multimillion dollar litigations where we really were helping more on the reputation side of anything.

10:03

And how big did it get, like, by the time it sold?

11:02

Yeah, we, we sold for around 105 million cash, similar amount of equity and, and so sizable. And we were like the team wasn't huge, but we did have pretty deep penetration across the Fortune 500.

11:04

So you're working at the acquirer when like ChatGPT happens, end of 2022. What does that do for you as a, as a founder who I have to assume was like thinking of going back in?

11:17

Yeah, for sure. I mean, look, I think part of it was I never would have like applied for the job to be leading AI strategy for this like big kind of software company. Right. But actually people like Kevin Scott, the CJ of Microsoft, is on the board of Relativity. Relativity has built an incredible business. And it was like super, super, like an old company, like 20 years old. But you know, they have really deep, like distribution and a lot of things that they've done really, really well. And so I think I just learned a lot from that experience. But at the same time, to your point, when a transformative technology comes out, like we saw with ChatGPT, and suddenly the market was like changed overnight, I just sort of couldn't help but think about this thesis that I started forming and really kind of recognizing that my own DNA is much more of a builder at heart.

11:27

Tell me more about the original thesis.

12:10

Yeah, I mean, look, I think a lot of it came down to this recognition that I think, and I want to be like, very clear because like, some things I'm going to say might sound a bit controversial around AI is the future of labor. And you read the manifesto on our website, like we're, we're pretty direct that we don't think this is like a tool or you know, a productivity assistant. Like the writing was on the wall to me that actually most knowledge work and the time and cost associated with it is asymptotically approaching zero. Which doesn't mean that humans don't have new jobs to do. And, and, and I have a lot to say about that, but ultimately like $6 trillion of professional services labor and you know, another 10 trillion if you count the internal teams are basically doing work that may today still be necessary, but tomorrow won't be necessary. And so like, the implications of that are pretty massive around. What does that actually imply? I mean, we're, we're, we're seeing in the last week now what's happening in the SaaS world, where suddenly everyone's calling a lot of things into question. But I think like at its core, the idea that we have a trillion dollars of like the law firm industry, which is basically packaging judgment into time and selling it by the hour, when now that judgment is actually should be infinitely accessible by people, just struck me as not only a huge business opportunity, but also from a mission standpoint, like 90% of people who need a lawyer right now can't afford one. And every small business, like we even paid a hundred thousand dollars to our law firms for our series A legal documents, right? And I now know that work can be done for a thousand dollars. So it's kind of like crazy to me that if you think about like the 2.5% of our GDP that's basically going to paying some kind of legal tax, like if we actually shifted that into paying for more innovation, engineering, things that are going to actually contribute like to something that's much more significant for the economy and for the world. That was a very, very exciting kind of opportunity to me.

12:12

Do you think? I mean, right now the world is very focused on what AI was going to do to the SaaS companies, the company selling software. Are you saying that they're not as much at risk, but it's the professional service that are, or it's all of them that are like equally at risk?

13:58

Yeah, great, great question. I mean, I would say it's personally even more so the services companies that are at risk. And I think both are at risk in like a similar way, which is, I don't think they're going to like die and go away, but I do think like you look at what's happening in the public markets right now, you know, the value of a company is basically you're looking at the terminal value, the growth rate, the time value of money. And so it's not like this, this big dip that we've seen post Anthropics release a week ago wasn't saying that a company like Salesforce has no value. It's just that the Historic growth rates that we've all projected into the future may no longer be as accurate. And the reality is companies that grow less quickly end up trading at like, much, much, much lower multiples. So the overall value goes down a lot. So I think in SaaS, people are frankly maybe even overreacting because they don't realize like how hard it is to get the kind of distribution. And as long as those companies can reinvent themselves, which I don't think is like impossible, I think it actually has to happen. It says more about how quickly can they do it. And if they can do it, I think they're going to be relatively better off. I think on the services side though, you know, to me you look at like the fact that historically we've deliberately made like to even become a lawyer, you have to pass the bar. There's so much regulation around it. And then the assumption is if you're a law student who goes to law school and you're now racking up 150k in debt, like you almost have to get 160k type of job coming out of law school. And like that whole model is basically predicated on again, this credentialed based, it's

14:12

artificial supply in a way.

15:38

Exactly. And now that AI really, like every person I know in the world, every business person, whether their lawyers like it or not, they are using AI to do their legal work. Right. And they're actually seeing that it's getting like at a minimum 90% of the way there. And so then they can still use a law firm, but it's going to be a much lower amount of work the law firm does. Right. Compared to before. And so that trend is going to continue. And I just think it's a matter of time before the whole notion of a law firm as we know it is going to be completely like transformed.

15:40

This is not like standard PMF show stuff. But because you spend a lot of time thinking about it, I'll just, I'll just keep going down this thread. Like you kind of alluded to this before. Like, are you kind of like, okay, we need like universal basic income sort of thing. There's going to be like no jobs. Or do you think that ultimately this is like any other technology? Yeah, there's going to be crazy ups and downs throughout. Maybe the number of people that work in law becomes like 10% as much, but something else comes in and easily takes over the other 90%. Like where are you on kind of that maybe more philosophical, like unknowable debate.

16:04

Yeah, for sure. So I actually hate Universal Basic Income, and I don't think that's like the right answer at all. And I do think, for what it's worth, like, no one has a crystal ball, so I have no evidence of this, but my own view is we're going to have as many or more jobs as we've ever had, and it's just going to require some like, structural resetting of many of these institutions in our, in our economy. And even our education system needs to change massively as well. But, like, the reason I sort of don't like Universal Basic Income, I can just start there, is it feels extremely like un American and sort of like the opposite of all the benefits of capitalism. There's obviously some downsides of capitalism, but like, at a very basic level, I think humans respond strongly to incentives. And I also think mastery, working hard, fighting some kind of pain and overcoming that is actually something that inherently is like, wired into us to make us much more fulfilled and happy as a species. So that, that's a really important mechanism to maintain. And UBI is kind of saying, hey, unfortunately, there's nothing for you to do anymore, so please take some money and watch, watch a TV show, right? Which, which I think is like, totally wrong. And when you look at like, let's just say with like product management, for example, right, you used to have the UX designer, the, the UI guy, the product manager, the engineer, maybe the front end, the back end engineer. And so it's kind of like this whole pod to be able to build a product. And now with AI, like one person who has the ability to navigate AI correctly can actually do the work of five, let's just say, right? And so to me, the implication of that is not great. The other four people get UBI now, right? It's actually maybe all five of them form a single unit and go out and create 5x the number of products. And I think we're already seeing that like, like everyone in the Valley, or really any tech company at this point is now using maybe previously cursed or increasingly Claude or whatever. The mechanism is right, to no longer write any code manually. But we're not hiring fewer developers, we're hiring way more. And so the whole Jevons paradox thing I actually do believe in, but I also think there's going to be a lot of new jobs that we're just seeing the beginning of what these look like, but they haven't existed historically. And as long as you have the right framework, I think the most simple one being you can make A ton of money. If you actually figure out a way to tap into some kind of economic value that was previously inaccessible, that that's a much better mechanism to me to keep people happy, to build and keep our economy growing, as opposed to just assuming that we're all, we're all done.

16:30

I think that makes a lot of sense. I mean, the UBI thing does sound like a top down solution to what's probably a bottoms up reality, more organic kind of thing. Like, it just reminds me of, you know, the issue with overpopulation. That was a big thing like 50 years ago and it was like some countries really had policies around limiting the number of births. And it turns out that this stuff just kind of takes care of itself because the incentives change as countries develop and there's education, all these other things that come together and that actually ended up being the answer. And it feels like, I mean, the history of mankind or of humans at least is like more aligned to the idea that yes, there's going to be structural disruption, but things realign and readjust and ultimately they get figured out and if anything they end up better than they used to be. So, I mean, we'll see how it all happens. But I'm probably more on that side of the fence than the other, maybe going back to the storyline. So you're seeing this, you think it's going to affect labor. Do you have like a crisp idea of what it is you want to build at that time?

18:53

Yeah, I mean, I think yes and no. I think I knew what I wanted to do a lot of discovery on. Right. Which were basically two categories of discovery that I was really eager to. And when I left Relativity, I spent the next few months both interviewing 80 chief legal officers, like before even writing any code and figuring out the product. But also, and you know, like the interviews were centered on maybe two primary topics. One was kind of the, not only the why, but the why now because I find that as a founder, probably my biggest learning from three startups in 15 years of doing this is that market timing is the most important factor in your success and the one that's the least in your control. So I really genuinely wanted to like discover without even biasing the person I was talking to, hey, what are your top three priorities and why? And, and frankly, I was, I was kind of suspicious that a lot of the chat GBT stuff at that time was actually just going to be hype that was going to die. Because the reality is the legal markets never change, like through all of history. Right. And so why would it suddenly let AI eat it on some level was kind of my number one topic on the GTM side, validating the market timing. And that was when I was shocked because even three years ago, it was suddenly becoming a top three priority for the chief Legal officer to like, figure out AI. Because post ChatGPT, it sent ripples through the entire, you know, like every boardroom. And the reality is law is a industry of generating language and analyzing language. And so the amount of exposure to automation by AI was just like, objectively maybe top of the list, right next to marketing or something. So that was like, super interesting. And then from a product perspective, I honestly just wanted to like, given that, like, the much easier path, I would argue is just to build another SaaS company. And the reality is 2% of the budget is still enough for a startup to build a big business. But like, I was very curious, what are the 98% of the, like the humans that are consuming 98% of the budget actually doing? Right. And I. And so we actually spent a lot of time first, like deeply, almost following around, like these lawyers and understanding what are they doing. And we looked at four areas, litigation, compliance, M and A and contracting. And tried to just figure out, like, the human side. Forget the tech. I obviously know what the tech's doing. Cause I've been in that world for a while. But what are the humans doing and where does the tech fall short?

19:50

How'd you do this, by the way? Like, logistically, did you like, go into their office, spend a day with them?

21:59

Good question. So actually, and this is maybe more on topic for the show, which is like Both with Text IQ, my last company, and with UDM, my current company, we've gotten 90% of our customers from cold emails that I've sent. So people always assume, oh, you sold your company, you must have all these connections. And you know, we had like one or two general counsels that did sign up for this for my last company. But the reality is my last company, we weren't selling at the C suite level. We were selling like two levels down. So most of those people weren't that relevant, at least to get to the chief legal officer, which is now our core icp. But, you know, I, I basically contacted a few and it was like, super interesting because initially it was like, hey, I'm a founder at an exit. I really believe AI is going to impact labor. I'm just really, really curious, like, to see whether you'd be open to sharing, you know, kind of some feedback on a 15 minute call that kind of email. And then I got a few people, like, I look at like Rob Beard, who at the time he had been the chief legal officer of Micron. He was just leaving to be the chief legal officer of MasterCard and one thing led to another and he was basically like, look, if, if you want to just like come in and see what we do, like, even if you don't have a product yet, I'm happy to just give you that opportunity to

22:03

cold email him or how did you get him?

23:09

Cold. Cold email.

23:10

Wild.

23:12

Yeah, almost all of them are cold emails. And yeah. And then so we, we basically had like, you know, five design partners initially who were chief legal officers that were like, they could see that we were pretty credible as a team and our technology background was real. And to be fair, my co founder, she had already built a lot of tech because he had no idea about legal, but he was basically like, all knowledge work basically entails very, very similar set of activities. So he'd already built an initial platform and we, we just said like, very honestly, we have no idea how this can help in legal. We want to first understand, like, where the time is going, where the money's going, where the bottlenecks are, how we can, like, what the actual technical problems are that we can solve. And I think like the biggest thing that we discovered was that it wasn't really an AI problem. Or maybe, to put it differently, every AI problem is actually a data problem. And the data problem was really the most interesting thing that we learned from these Fortune 500 companies is all of the knowledge that they need. Whether it's a human who joins their team or an AI agent who wants to join their team, that knowledge is currently distributed across a variety of SaaS applications and trapped in people's heads. So like the real pain was forget about AI. Like when my colleague goes on paternity leave, it's a nightmare because they know how we negotiate contracts with this type of customer. And unfortunately no one's ever codified that knowledge. So we actually started to realize like, that's the real problem. And could we build a platform that actually, frankly is more of a data and knowledge platform? Of course AI becomes like a core component of it. But, but AI is almost like the last step, the first step is how do you actually solve the data problem? And once we started to do that, even in a small scale way, we were getting like incredible performance and that, that became actually the big differentiator.

23:12

I'm going to ask you for a small favor, a tiny Little favor. In fact, it's not even now that I think about it, it's not even really a favor for me. I'm actually trying to help you do a favor for you. Just hit the follow button. You won't miss out on the next episode. You'll see everything that we release. If you don't want to listen to an episode, you just skip it. But at least you don't miss out. Let's go deeper on this setup with the design partners where you're finding the ultimate problem, which ends up being this data problem. Because everybody knows these days, because a lean startup and all that, that you have to talk to customers. You know, design partner is like the new beta customer. And frankly, just about everybody is doing it. What I find and when I talk to founders who are like in the game doing it right now is it's not equally executed, like everyone's doing it, but not at the same level. And ultimately product market fit. One of the biggest ingredients is whether you end up picking something that really matters and you find the way you really should solve it. I mean, if you get that right, everything else starts to take care of itself. And if you don't, it's kind of like you're pushing a boulder up the hill the whole time. So how are you setting up? You got these five design partners, like, what are you doing with them? To see the sort of things that you see and all these little details and the subtleties that ends up, you know, giving you this kind of a ha.

24:45

Moment around.

26:01

Around data and then. And then you go and solve that.

26:02

Yeah, good question. So, so look, I, I think there's also like founder market fit, right? And it's really important that you're honest with yourself about where are you likely to crush it versus not be really differentiated. And I would say one of my learnings is because of my last company's experience, I, I learned I actually a just happened to like relate pretty well to the C suite people. I don't know. This just, just kind of inherently like what I found with my last company is sometimes it was like the junior person who's like day to day, deep in the weeds. We did really well with. And then actually when we like skipped one or two levels and went to the top, they also really loved what we were doing. We had the most challenges with kind of the middle management layer who's like, not really. Actually close enough to the workflow or the day to day, but also not really an executive who's empowered to make decisions. And so One I just knew like from my perspective with enterprise software and part of why I love B2B versus consumer is like there's actually a lot of repeatability. Like if I were to I could literally probably draw on a piece of paper like the far like the five archetypes of a chief legal officer and within like 10 minutes of meeting one now I can tell you whether or not they like prefer to have bagel and mocks for breakfast or go to vacation in Maine versus Colorado. Like it's like very, very like psychological profile is actually very very similar. And it's like a finite number of archetypes I would say which once you figure out, okay, we now know the five profiles like across the Global 2000 we can actually clean up because this is like very, very consistent. I would say so, so one, like I started to recognize, okay, I actually do quite well and understand like viscerally how these executives think because for them it's actually all about like value. And at that level they'll find money to pay for something if it's adding a ton of value, right? But it's very, very rare to actually find something that does add value and rare to find someone who actually understands them like deeply. Right? So for me what we did and this kind of mirrored with like my own journey as a founder, which is like when I look back on it, I think most first time founders or founders who haven't yet had a good exit, deep down they're often playing not to lose, they're not playing to win because the reality is you put so much into a startup, it's your whole life, you're making very little money compared to what you could be making. And like at some point you're just like, this is so painful. Like I just need to get some kind of return on this investment. Especially after years of pain and suffering.

26:04

Which by the way makes, it makes a lot of sense that you think about owning 30% of a $20 million exit. $30 million exit, let's say that's $10 million. Like that's life changing money, right? So you can't, you can't blame them a hundred percent.

28:26

I don't blame them at all and I totally get it. But I think just to like reflect on my own difference this time around, like I know I've never felt more aligned with our VCs because this time it's like if I'm going to put all this like my entire life into something, I really want to play to win this time, not play not to Lose. And so even back then, like two and a half years ago, I went to each of the design partners, I told the chief legal officer, and I insisted it had to be their chief legal officer, not their deputy. Right? Like, so reporting the CEO of these Fortune 500 companies, I basically said, hey, look, before we even start doing anything together because your time's so valuable, you should just know we're only working with companies where we see an opportunity to deliver a hundred million in value in five years. At least a hundred million dollars in value in five years. And so if, if you're more talking to me because you want to explore whether or not ChatGPT can summarize NDAs for you, like, literally there's a thousand legal AI companies I can point you to, maybe that time, 500 now. A thousand, right, who have like a point solution that's probably really good at having a feature to do some small thing. It'll probably actually charge them very, very little. But I was like, if we're aligned that the goal here is to unlock a hundred million dollar plus of value, then this entire design partnership is not only proving that our technology works, but also like mutually uncovering. Where does that value actually exist? Right. And so that was like, I think one, one big difference. And it meant that by the time we kind of actually found the right initial pilot, everything was with that frame of how do we scale this as quickly as possible? How do we actually build a platform, not a product, and start tackling like multiple use cases and even expanding beyond legal as quickly as possible.

28:37

But what did you do with those design partners? Did you go into the office every day? Did you just have a call with them? Did you build stuff and show it to them? And that was the iteration cycle?

30:05

Yeah, yeah. I mean, look, I, I think part of it was realizing the vision that they had more than we had. So I think one of the learnings was, you know, for the chief legal officer, they think about risk very differently than everybody else in the organization. And they also think about like, for example, contracts. They think about them on like a portfolio level in a way where today most lawyers almost act like every time they see a contract, they've never seen a contract before, and they kind of repeat the same workflow. And so we started to get like a better vision of, okay, almost like whiteboarding. In your ideal world, you'd want to be able to see like, you, your portfolio of contracts. And if there's, you know, 5,000 of them, you want to know, like, what's the one that's like the highest risk where I have to change one thing or some data to help me understand that. The last 10 times I tried to get limitation of liability under 10 million, I never got it. So I can try again, but if I do, it's going to add six weeks to the sales cycle and my CEO is going to hate me or I can just concede that term. So kind of like we started to understand like what data do they actually need to be able to make better decisions about risk fundamentally and those trade offs between growing the business and protecting the company. And then once we had that, we started to go like really deep into the weeds with like let me just spend all of our time with like the one person who's actually doing these contracts today and starting to build again like initially more of a co pilot and then trying to figure out where can the AI become pilot and where does a human become pilot as opposed to just being surface level about that and then starting to build like real case studies where we can now show a small scale. We've gotten like 55% efficiency gain, 93% F1 score. We're eliminating the need to spend $4 million a year on a law firm for this one use case. And that justifies the million dollars to udf for example.

30:13

What was like one of the first use cases that you solved that was, that was meaningful?

31:50

Yeah, for sure. So I think like two of the most common ones, one was around contracting, but not like CLM or contract lifecycle management, which is kind of an established and it might be a little bit of an old category. So instead of like trying to build the software to put all your contracts into one repository, we started to think a lot more about two things. One, if we wanted to like eliminate 90% of the human review, what would be required to do that? And so that was like one category where we really developed really cool technology to be able to do that. The other was around like compliance. And specifically a lot of these in house teams do like marketing reviews. And we were able to basically like learn that AI out of the box will never be that amazing to somebody who actually knows what they're doing. It's amazing to somebody who has no idea what they're doing. But to SME they're like, this is crap. And so we almost learned like, okay, assume it's going to be crap in the beginning. And our job is not to have the first output blow your mind, it's actually to have the third output blow your mind. And in the course of from One to three. What we're really doing is basically leveraging our data platform to try to capture all this institutional knowledge that's trapped in your head as well as look at like, what are all the past examples, all the data that we can get our hands on that helps train the model and how you customer X actually think about risk. And that was like a really effective way where again, like within three weeks we were seeing these like crazy performance improvements from like 50 to 95% F1 score and efficiency gains that were very palpable. But even more importantly, subjectively, like, lawyers are actually very subjective around things like language. Right. So like we had a customer where the chief legal officer just hates adverbs. Adverbs aren't legally right or wrong, but the brain that we built for that customer would never use an adverb. And they can tell, like, they're like, oh, wow, it's starting to how I actually do my legal reviews. And so the model's being trained off of my own, my own data.

31:54

Were you funded during this time or like, did you just kind of bootstrap up to a certain point?

33:45

Yeah, we. We raised a small, like. Well, not small by all standards, but like a $6 million seed round. That was Mike Maples, who's a longtime mentor and back to my last company. You know, he. He basically not only funded this when it was like an idea on a napkin, but. But also I met him like every week to kind of like, he was almost like a silent co founder in terms of the value that he. He added in the beginning too.

33:49

How long did that last before you raised the. The A hundred million dollars round?

34:12

Yeah, that was about nine months. So not too long, but long enough that we had like, our seed round was like. It was a very, very weird pitch deck, if you can even call it that.

34:15

It was like, when was this, by the way? End of 20N23.

34:25

Yeah. And it was literally here's like an 80 page transcript of all the interviews that I did with the buyers and then obviously used AI to help surface some of the key insights and all that. But that was our pitch deck. It was like we've now validated at a very deep level that this buyer is ready to go. 90% of them are making AI a top three priority. We then like, as soon as we closed around, I flew to Copenhagen for our first design partner and we actually closed our first 250k paid year kind of ARR. Because we'd really kind of already built something that was useful and that's like stepping back, I think this, this approach that we've taken, like honestly we also got kind of lucky in terms of all the conventional wisdom of like build an mvp, start really small. Like none of that makes sense anymore when we can all build so quickly. Now with Genai, it's more of a race of who can actually build something useful the fastest as opposed to like starting really small. Right.

34:29

The sort of things that you're building for these in house legal teams, are they similar to the sort of things that Harvey and Ligor are building for the law firms?

35:18

Yeah, great question. So I, I think again I'm sure Harvey and Lagora are now they almost have to move into the enterprise space because the law firm market's only so big. But you know, their, their core, I would say product was built for law firms. And I think who your product's built for initially, like it's very, very hard to get away from like even your mindset, right. Where people talk about legal attack implying that like as, as long as you're a lawyer you should have the same product. But in my experience, law firms need like totally different things than enterprises and we don't really see ourselves as a legal tech company. The last six months especially most of our growth is actually coming from procurement and compliance now. So once we proved to the chief legal officer that it worked in legal, oftentimes the same buyer actually oversees procurement and compliance. It wasn't even going to someone new. It was just like, great, let's take the other 30% of my budget and start applying it there. Right. And so I think the probably biggest difference is everything we've done has been from an in house perspective. And so when we drive outcomes for our customers that could be reducing my law firm bill or alliance, it could be speeding up contracting review, it could be helping you with compliance or procurement type of use cases. And I'd say like some of that has overlap with what a law firm would need. Like legal research for example, is pretty universally the same set of laws for example, or things like litigation and M and A are probably because the in house team shifts it out to a law firm. Those products have some overlap, but I think most of the rest of it honestly is actually quite different.

35:25

And then once you close this 250k ARR, are you like selling to 3, 4, 5 others or are you focusing because, because it's enterprise on getting the product to a certain level with this one customer?

36:50

Yeah, so, so we're, we're pretty unusual and thankfully I had the Board support on this. But what I've seen is way more startups die from prematurely scaling than from really deeply understanding what they can uniquely provide that their customer is desperate for. So we spent like, you know, up until September of this past year, 25, we were literally still heads down. First year only 10 customers. Second year we expanded it to 20. And then in September of this past year we were like, okay, we're now seeing across all 20, a lot of repeatability. And it's basically the same platform and the same solutions that are now scaling without us putting in heroic effort and all that to be able to make it work for each customer. And then we kind of turned on the scaling motion and went from 20 to 60 pretty quickly.

37:00

How many people were you when you, when you closed that 250k?

37:42

4.

37:46

Okay. Tiny.

37:48

Yeah.

37:50

And then how much did you raise? That was your seed with 6 million. You said like halfway through 24 you raised an A. Is that right? How much was that round?

37:50

So we, we did it in sort of two tranches. And, and yeah, I mean it was, it was more like towards the end of 24 and then we did it in two tranches. So the first tranche was like 30 million more for just purely supporting the organic growth. And then we had this thesis around using M and a couple of creatively because we met over a hundred law firms, legal services companies, and we were basically trying to figure out, could any of these, like potentially be a way for us to accelerate data access trust with a set of customers where they'd already been doing a lot of their legal work. And if they now shifted the model from hourly to output based, we could sort of turn on our AI and actually more quickly give the customer kind of the best of both worlds where they're already getting a trusted human they already work with. And now we're also, from a product perspective, starting to really learn much more deeply again to really transform the labor part of it, to offer something that, that's truly an alternative to kind of what they're getting from, from people or law firms today.

37:56

Yeah, maybe walk me through it because I know you ended up buying a law firm and you know, adding AI to it to kind of deliver outcome based law. How does that fit in? It's not immediately clear. Like you're right now you're working with, let's say at that point in the story, right? You're working with chief legal officers in house legal teams. You're building a platform for them to drive productivity and efficiency and effectively not need to do as much manual work with whatever, like legal stuff they're doing, contracting, etc. Why do you need to buy a law firm? Like where does that fit into the kind of strategy?

38:53

Yeah, yeah, so it wasn't a law firm, it's a legal services company. They have legal professionals who are acting as an extension of the in house teams of OpenAI, Stripe, Citibank. And so this is again kind of like unique to this world, but a lot of in house legal teams, like you have three choices as a chief legal officer, you either hire a full time employee you give to work to a law firm, or there's kind of this like third category that a lot of people don't know about, which is like alternative legal service providers, ALSPs. And the idea there is like axiom is a very well known one where essentially they're like often former lawyers who practiced either in house or at law firms who decided to kind of go down this third path where basically they do for example a lot of contract review and negotiation work. They offer much more attractive prices than law firms, but they can also get like pretty deeply embedded with like one or very, very small number of clients where they start to get to know everything about how OpenAI likes to do their contracting. And so we bought a company who was like that third category.

39:21

But they're not lawyers, they're like paralegals. They're kind of operating in this weird

40:28

place or if people are usually lawyers, you know, from a regulatory standpoint, just there's a lot of nuance in like okay, calling something a law firm versus not. So technically they're not a law firm and technically they're not giving legal advice. Right. But they will be 90% lawyers or people with a legal background who have earned the trust and are delivering really good quality outputs. Right. To a lot of those in house legal teams on their day to day work.

40:31

So you buy one of these. What's, what's the strategy behind it?

40:57

Yeah, basically we just wanted to test, you know, kind of three main things. Number one, could we more quickly start working with their customers because they already have a relationship, they're already a vendor, they've already gone through infosec, etc. Secondly, could we start expanding share of wallet with our customers? Because the reality is as a tech company, we had no idea how to actually like offer experienced contracting professionals. And AI is not going to like take over that work on day one. It's going to be a bit of a journey, right, to go from humans doing it manually to kind of AI out of the box. Like there is a bridge that you have to kind of cross there. Right. So could we basically expand, share a wallet with our customers? And then the third is from a product development standpoint, our hypothesis was that by owning the humans who actually do this work, we would be able to go way deeper to actually understand how do you convert services into software and start really figuring out like how to offer like not, not just, you know, faster and cheaper, but actually better. Right. And I think all three really got proven out pretty quickly in terms of we were suddenly signing up their customers at sales cycles that were like 50 to 80% faster than if we'd gone to one of those customers on our own.

41:00

And you're selling them the in house AI to their customers?

42:11

Yeah, I mean we're selling them two things. One is the in house AI, so a whole new set of products that we developed. But then also even for their existing work, like you think about a customer who's saying, okay, great, I'll take two of these legal professionals to help me negotiate and review my contracts. What we did is say, hey, instead of paying us by the hour or by the number of humans, we are now changing our pricing model. So we want you to pay for the amount of contracts we review, for example, or a fixed fee. So now our incentives are now aligned to use technology. And once we did that, we were able to say good news, Instead of waiting five hours, you're now getting the contract back and 30 minutes. So the turnaround time's way faster. And on our end, Instead of a 30% gross margin, we can see an 80% gross margin because we're not using AI to do most of the work.

42:14

And then walking through one thing like why not just go all in on that? Like, but that just seems like such a no brainer value proposition to go to any enterprise anywhere in the world and say you're paying X to get these contracts, you're just going to pay like half of X and you're going to get the contracts twice as fast. It's almost like where do I sign? Do you know what I mean? Like, why was that not the thing versus like what you're, what you're mainly doing.

42:58

Yeah, great, great question. And, and this is again like, it all comes down to like the strategy of each company. I think for what it's worth, since we launched our AI native law firm, there's been like a dozen people who are now doing this and it's obviously becoming like a more and more crowded space. My Honest answer is I think a lot more about how do we get to a billion in revenue by 2030 and build a generational company and win this market than I do. How do we like optimize for sales cycles and this year's revenue? Right. And I think it's a subtle but important distinction because when you look at like most of the work people are farming out to these kinds of companies and even law firms, it's not like super strategic complex. It's more like, I believe I still need a lawyer to do this and so I'm going to use a lawyer to do this. But, but like in practice when you step back, a lot of that work can be completely automated over time with the right platform. And so we're much less interested in some near term opportunity to capture revenue from other services companies and you know, kind of temporarily gain some advantage there, and much more interested in like, how do we actually build this brain that over time really deeply understands at a more strategic level how to think about risk. And we, we were seeing so much traction on the organic side that it was like impossible to ignore that. Right. Like the companies wanted to buy our products. Like we almost couldn't build things quickly enough for them to be able to consume more of our technology. And meanwhile we had this like proven but ultimately kind of limited opportunity in the long run to take over a lot of the services spend as well.

43:20

Perfect. Well, let me stop there. Let me ask a few questions that we, that we typically end on. First one is how fast did you hit a million ARR from the time you started selling?

44:55

Yeah, we hit a million ARR right around six months.

45:03

And, and then what's like growth been since, like from the 1 to 10 range?

45:05

Yeah, we went from roughly because, because again, for the first couple years I would say ARR was not what we were optimizing for. It was actually still, I believe each customer will become a 10 million ARR customer for UDIA because we're driving $100 million in value over time. And so that hasn't changed. And if anything, it's gotten more the case as we expand beyond legal into other back office functions. Right. Where there's just a lot of people, internal and external, doing work that can now be mostly automated. So we weren't focused on ARR, I would say for the first couple years, but in the last 12 months we went from 2 to 20 million ARR.

45:10

When was the moment when you felt like you'd found true product market fit?

45:45

Yeah, I think it was probably like right around last summer, where we still are now seeing like 90% of our business comes from customer referrals. But when you hear any of our customers talk about Udia and we haven't done, like, we haven't invested a lot in marketing, we're not that well known. I would say we're more under the radar still. But, like, that's how I started to feel like we have product market fit is anytime anyone talks to one of our customers, we get another customer. So, like, our whole GTM engine is kind of how do we just like, try to get our current customers in a room with prospects and at that point, like, we can walk away, which I think was a pretty good sign that we're onto something.

45:48

Things sound like they've been like, on a tear basically from the beginning. I mean, even though, like you said, ARR wasn't the focus, it sounds like things have just been up and to the right. But was there ever a time where you actually doubted if things would really work out?

46:26

Oh, yeah, yeah. And I mean, you know, frankly, like, I think even now it's hard to feel good about 20 million ARR when you see everyone else in the world getting to 100 million arrangements in 12 months. You know what I mean? And I think personally, I think ARR this would be a very unpopular view, but I don't think it's actually the right metric, like the single right metric to focus on anymore. Because there's a lot of examples with these LLM wrapper companies where it's like, temporarily, you can appear to be magic for a lot of the market who either doesn't have the technical chops to go directly to the foundation model, or the foundation model hasn't yet moved deep enough into the application layer for someone to see that actually that's a viable alternative that can do the same thing at like one tenth the cost or 100th the cost or whatever it is. Right. So I, I, I think it's, it's been, it's still been hard because the reality is like, we're all human. And when you see like a bunch of companies that everyone's heard of reaching even way more significant revenue milestones much more quickly, it's like part of these, like, what the hell are we just kind of like doing the wrong thing here? But I think being more experienced founder, I've, I've had to remind myself, like, our thesis really is if we go really deep, if we build trust with these C suite buyers, and ultimately it's their data and their dollars that are flowing through the ecosystem, then there is a lot of value. Even if it takes longer to get ourselves like deeply embedded into the workflows and getting access to the data, that is actually pretty proprietary and institutional knowledge. We know that's the right strategy, but often we question, because we just aren't growing as fast as other companies in the space, whether or not it is the right strategy. But we always come out with like, no, let's just stay focused on ourselves and ignore competition and stay focused on the customer for the most part.

46:38

And then, last question. What would be like a top piece of advice that you would have for an early stage founder that's still like looking for product market fit?

48:15

I think it's just really asking yourself that one question. What can you uniquely provide that your customer is desperate for? That's my definition of product market fit, which, which I didn't come up with. I, I stole from Mike Maple, so I think stole from someone else. But, but you know, it's, it's, it's really desperation, not just willingness to pay and then also what you can uniquely provide, which in this world of, you know, AI eating the world, increasingly gets called into question. But like in our case, like, trust with chief legal officers combined with like a data and knowledge platform were kind of our answer to that question. And then we learned what they're desperate for after spending a lot of time with them.

48:22

Perfect. Well, Lamar, man, thanks for jumping on the show, dude. It's been great.

48:59

Yeah, thank you for having me. Really, really good to be here.

49:02

Wow, what an episode. You're probably in awe. You're in absolute shock. You're like, that helped me so much. So guess what? Now it's your turn to help someone else. Share the episode in the WhatsApp group you have with founders. Share it on that Slack channel. Send it to your founder friends and help them out. Trust me, they will love you for it.

49:06